Structured abstract:
Introduction: This study examined the effects on drivers’ yields of seven conditions (pedestrian behaviors) at a roundabout, measuring yielding behaviors.
Methods: At a roundabout, drivers’ yielding responses were recorded (none, soft, hard) when a pedestrian presented one of seven conditions: pedestrian without long cane, long cane display, hand up, reversible step (a single step forward into the street), head turn gaze, hand up plus gaze, and a reversible step plus gaze. Trials were completed at the crosswalks at the entrance and exit lanes.
Results: At the entry and exit lanes, the hand up and reversible step conditions received significantly higher yield percentages than all other conditions. At the exit lanes, the reversible step condition received a significantly higher yield percentage than the hand up condition. There was no significant difference among the three head turn gaze conditions: hand up, hand up plus gaze, and reversible step plus gaze. The hand up and reversible step strategies averaged 83% yielding overall, compared with 20% yielding for the other strategies.
Discussion: This study demonstrates that forced detectable yields may be generated when pedestrians have not entered the travel lanes of the street at a crosswalk or have only entered the very edge of the street. Gazing behaviors do not substantially influence drivers’ yielding rates.
Implications for practitioners: Using a hand up or reversible step can allow pedestrians who are visually impaired or blind to create circumstances to help determine if and when to initiate a crossing at a roundabout.
Ashmead, Guth, Wall, Long, and Ponchillia (2005) explained that the behaviors and safety of pedestrians who are visually impaired at roundabouts were not well understood. To address these challenges, some orientation and mobility (O&M) specialists and traffic engineers have examined drivers’ and pedestrians’ behaviors at roundabouts. Drivers’ yielding behaviors have been reported in response to visually impaired pedestrians who were traveling with and without mobility devices (long cane or dog guide). Increased yielding by drivers can represent a substantially decreased level of risk of injury to pedestrians (Ashmead et al., 2005; Geruschat et al., 2011; Guth et al., 2005). O&M specialists and engineers have also looked at road treatments and signalization to reduce risks at roundabout crosswalks (Inman et al., 2006; Schroeder et al., 2011). No research to date has identified tools or strategies to allow visually impaired pedestrians to cross securely at roundabout crosswalks. In a review of modern roundabouts, Long, Guth, Ashmead, Wall Emerson, and Ponchillia (2005) surveyed the challenges to accessibility for pedestrians who are visually impaired, articulated the problem, and offered key questions:
- How may changes in driver-pedestrian interaction affect access to roundabouts?
- What role does movement by visually impaired pedestrians that may communicate to drivers their intention to cross play in increasing yielding behavior by drivers?
- Can visually impaired pedestrians detect yielded vehicles reliably? and
- If drivers yield reliably to pedestrians
and pedestrians can readily detect when
drivers have yielded, then access to
roundabouts will presumably be improved (p. 620).
Literature review
YIELDING AND PEDESTRIAN BEHAVIORS
At roundabouts, Ashmead et al. (2005) noted that drivers yielded more frequently at entrances than at exits and, when compared to their sighted peers, participants who are visually impaired required more safety interventions in their attempts to cross intersections at roundabouts and they rarely took advantage of drivers’ yielding. The researchers pointed out the information limitations for auditory-only traffic monitoring and, further, that drivers’ yielding was often not useful for pedestrians who are visually impaired. Geruschat and Hassan (2005), in a study at two roundabouts, found that yielding increased from 52% to 63% when a pedestrian held a long cane, and that an increase in yielding was greater at locations where yielding was unlikely. And Guth et al. (2005) found that the use of a mobility device increased drivers’ yielding, writing “[a]t the roundabout’s exit lane, for example, yielding increased from 4% with no dog or cane to 21% and 29% for the dog and cane conditions, respectively” (p. 327). This low level of yielding was not considered sufficient to make a substantial difference to pedestrians who are visually impaired.
GAZING AND HEAD MOVEMENT AT ROUNDABOUTS
There has been limited study at roundabouts of the gaze and head movements of pedestrians who are or appear to be visually impaired. Geruschat, Hassan, and Turano (2003, 2005) examined visual impairment and head movements at crosswalks, including at roundabouts. They found that a majority of pedestrians turned their head in the direction of the oncoming vehicles in the final seconds before crossing. There was no statistically significant difference between sighted and visually impaired participants; however, the pedestrians who are visually impaired did so less frequently. Those authors noted that head turning could be necessary for visually impaired pedestrians to collect information that is critical to safe crossings and defined this behavior as “consistent with the safe street-crossing” (p. 23). They recommended O&M training that emphasized effective head movements. It is worth noting that the authors addressed pedestrians with low vision, not those who are blind.
PROCEDURE
At the roundabout, a collaborating male pedestrian, six feet tall, wore dark casual clothing, a cap, and dark sunglasses. An engineering traffic study for the annual average daily traffic (AADT) of vehicles approaching from the west was 1,727 (Kalamazoo Area Transportation Study, 2011). A review of video data found that, during the study trials, vehicular speed averaged about 15 to 20 miles per hour at the entrance and exit lanes (when drivers were not yielding); during data collection the traffic could be characterized as low to moderate in volume, with substantial gaps. The entry crossing was a single lane, and the exit lane crossing was two lanes.
The pedestrian started each trial by standing less than one foot from the edge of the street at the crosswalk. The behaviors assessed included:
- Pedestrian without long cane: The pedestrian stood on the sidewalk, looking forward.
- Long cane display: The pedestrian held a long white cane at the curb so that it was visible to drivers in approaching vehicles.
- Hand up:
While displaying a long white cane, the pedestrian held a hand up with the palm facing an approaching vehicle. - Reversible step: The pedestrian took a single step off the sidewalk into the gutter of the street while flagging the cane, swinging it from side to side twice to waist level.
- Head turn gaze: While displaying a long white cane, the pedestrian turned his head toward the potentially turning vehicle and held his gaze at the vehicle or vehicles.
- Hand up plus gaze: While displaying a long white cane, the pedestrian combined the hand up with the head turn and gaze toward the approaching vehicle or vehicles.
- Reversible step plus gaze: Using a long white cane, the pedestrian combined the reversible step with the head turn and gaze toward the approaching vehicle or vehicles.
Data were collected in two sessions. The first collection used the conditions: pedestrian without cane, cane display, hand up (see Figure 1), reversible step (see Figure 2), and head turn gaze at the roundabout entrance and exit; the second used the conditions: hand up, hand up plus gaze, and reversible step plus gaze at the exit only. The combined conditions were all initiated when an approaching vehicle on the entry or the exit lane was approximately 130 feet away from the crosswalk. The collaborating pedestrian could nearly always hear the approaching vehicles at this distance, but for consistency another researcher verbally cued the experimenter when a car was at the proscribed distance. Trials were begun when there were no vehicles approaching the roundabout whose drivers could see the pedestrian approach the crossing.
Data collected included a judgment of whether the approaching vehicle yielded for the pedestrian, as has been done in other research on yielding to pedestrians who are visually impaired at roundabout crossings. Judgments of yields were made by three experimenters to allow for reliability of coding to be calculated. All three experimenters were experienced O&M specialists, and they based their judgments on observation of the drivers’ behaviors, the speed and deceleration of the vehicles, and other things that were happening in the environment.
The three experimenters coded yields at both the entry (n = 102) and exit (n = 100) lanes. Experimenters coded whether a vehicle yielded at all and, if it did, whether the yield was a hard or soft yield. Hard yields were defined as instances in which the driver was yielding for the pedestrian and the vehicle’s wheels stopped turning (for instance, the vehicle came to a full stop). Soft yields (also known as rolling yields) were defined as instances in which the driver was yielding for the pedestrian (for instance, the vehicle slowed appreciably or the driver motioned for the pedestrian to cross), but the wheels of the vehicle never stopped turning. Also, raters collected qualitative data about their observations and impressions.
One experimenter coded all trials, while the other two coders were used for reliability checking. Each of the secondary yield coders recorded 85% of all of the trials. The principal yield coder agreed with the first secondary coder on 90.7% of the trials, and agreed with the second secondary coder on 91.9% of the trials. When only yield versus no yield was considered, the primary rater agreed with the first and second secondary coders in 92.2% and 95.2% of the trials, respectively. Due to the high level of agreement, yields recorded by the principal yield coder were principally used in the analyses. However, comparison analyses were also conducted that coded a trial as having a yield if any of the three coders indicated a yield on that trial. Also, raters collected qualitative data about their observations and impressions.
Results
Table 1 shows the numbers of yields coded by the principal yield coder, as well as yield percentages, broken down by experimental condition, for entry and exit lanes at the roundabout.
Table 1 Yields and yield percentages by condition.
|
Condition |
Hard yields |
Soft yields |
No yields |
% yields (any yield, primary coder) |
| Entry lane | ||||
| Ped without cane | 1 | 5 | 14 | 30 |
| Cane display | 1 | 6 | 13 | 35 |
| Hand up | 15 | 6 | 0 | 100 |
| Rev step | 16 | 4 | 0 | 100 |
| Head turn gaze | 2 | 4 | 15 | 28.6 |
| Exit lane | ||||
| Ped without cane | 1 | 0 | 19 | 5 |
| Cane display | 0 | 2 | 18 | 10 |
| Hand up | 8 | 4 | 8 | 60 |
| Rev step | 10 | 9 | 1 | 95 |
|
Head turn gaze
|
1
|
1
|
18
|
10
|
The detailed analysis for each condition follows. When hard yields and soft yields were combined into one yielding category, there were significant differences across the five conditions in the yielding percentages at the entry lane as discovered by the primary coder, x2(4) = 48.18, p < 0001; as well as at the exit lane, X2(4) = 55.30, p < .0001. Results were the same when the data of any coder was analyzed. At the entry lane, the hand up and reversible step conditions received significantly higher yield percentages when either was compared to the pedestrian without a cane, X2(1) = 22.32, p < .0001 and X2(1) = 21.54, p < .0001, respectively; cane display, X2(1) = 19.99, p < .0001 and X2(1) = 19.26, p < .0001, respectively; or head turn gaze, X2(1) = 23.33, p < .0001 and X2(1) = 22.53, p < .0001, respectively.
At the exit lanes, the hand up condition received a significantly higher yield percentage than pedestrian without cane, X2(1) = 13.79, p < .001; cane display, X2(1) = 10.99, p = .001; or head turn gaze, X2(1) = 10.99, p = .001. The reversible step condition also received a significantly higher yield percentage than pedestrian without cane, X2(1) = 32.4, p < .001. In a difference from the entry lanes, at the exit lanes the reversible step condition received a significantly higher yield percentage than the hand up condition, X2(1) = 7.03, p = .008.
When hard yields were compared to soft yields at the entry lane, there was a significant difference across conditions, X2(4) = 16.76, p = .002. Further analysis of the data shows that the hand up and reversible step conditions had 71.4% and 80% hard yields, respectively (and no trials with no yield) while the other conditions had either 5% or 9.5% hard yields (with a high percentage of trials with no yield). This pattern of results indicates that the hand up and reversible step conditions not only garnered more yields, but a higher percentage of hard yields. A similar pattern of differences in hard and soft yields across conditions was observed at the exit lane crossing, but due to the high number of trials with no yield several conditions had so few trials with any yield that the Kruskal Wallis test was non- significant, X2(4) = 3.99, p = .41. The pedestrian without cane, cane display, and head turn gaze conditions had only one or two trials each with any yield at all, so patterns of soft versus hard yields could not be evaluated reliably. The hand up and reversible step conditions had slightly lower percentages of hard yields than seen at the entry lane crossing (40% for hand up and 50% for reversible step).
On a second data collection day, we replicated the hand up condition at the exit lane crossing to see whether this behavior resulted in a similarly low yield rate (60%) as on the previous day. In order to further investigate the effect of the head turn gaze condition, we also paired the head turn gaze condition with each of the hand up and reversible step conditions. The intention was to see whether adding the gaze behavior to either the hand up or reversible step could increase yielding for these two most promising yield-causing behaviors. Table 2 shows the yielding percentages for these three conditions (hand up, hand up with head turn gaze, and reversible step with head turn gaze) at the exit lane crossing.
Table 2 Yielding percentages at the exit lane crossing.
|
Condition |
Hard
yield |
Soft
yield |
No
yield |
Yield percentage
(any yield, primary coder) |
|
| Hand up | 10 | 3 | 7 | 65 | |
| Hand up plus gaze | 13 | 3 | 4 | 80 | |
| Reversible step plus gaze | 12 | 4 | 4 | 80 |
Table 3 Yield/no yield events by pedestrian behavior.
| Condition | Yield | No Yield |
| Raise hand or reversible step | 117 | 24 |
| Any of the other 3 strategies | 24 | 97 |
When displayed in this manner, the data clearly show the relative effectiveness of raising the hand or performing the reversible step. These two strategies are significantly more effective than the other strategies, X2(1) = 104.46, p < .0001, with the hand up and reversible step averaging 83% yielding, and the other strategies averaging 20% yielding.
Discussion and conclusions
Wall Emerson and Sauerburger (2008) noted: "The situations in which a pedestrian with a visual impairment is expected to cross without the advantage of a WALK signal are increasing, especially in light of the increased construction of roundabout intersections" (p. 757). They concluded that using the lack of traffic sounds for determining when it is clear to cross is no longer viable in all situations. For those situations in which it is not viable, one alternative might be to cross when drivers stop for the pedestrian. A forced yield, one that is created by the presence or action of a pedestrian when he or she has the right of way, can present such a crossing opportunity. Schroeder et al. (2011) said a forced yield ". . . occurs when a pedestrian steps out onto the roadway prior to the driver initiating the yielding process" (p. 44). However, the current study demonstrates that forced yields may be generated when pedestrians have not entered the crosswalk at all, or have only entered the very edge of the crosswalk, thereby mitigating the risk to the pedestrian from the approaching vehicle.
Any yield that represents a crossing opportunity must not only occur, but it must also be discernable by the pedestrian. The collaborating pedestrian attended to and reported the detectability of yields in qualitative data. Because vehicle speed is highly correlated to vehicle noise and detection (Man, Cho, & Choi, 2007; Wall Emerson, Naghshineh, Hapeman, & Wiener, 2011), slow yields could not be heard and, alternatively, all hard yields were readily detectable.
Two conditions offered statistically significant rates of yielding above the traditional cane display. They included the hand up and the reversible step. At the roundabout entrance lane (see Table 1) the hand up pedestrian behavior produced overall yielding of 100%, with greater than 74% hard yields; and at exit lanes the hand up resulted in overall yielding of 60%, with 67% of those being hard yields. The reversible step pedestrian behavior was even more substantial, markedly at the exit lane. At the roundabout entrance lane (see Table 1) overall yielding was 100%, with 80% hard yields; and in exit lanes the reversible step resulted in overall yielding of 95%, with 50% being hard yields (see Figure 3). It is worth noting that the pedestrian had the right-of-way to enter the crosswalk. The results represent not only considerable driver yielding responses, but present practical and useable results for pedestrians who are visually impaired at roundabouts with single lanes. However, pedestrians who are blind and visually impaired need to be aware that these strategies will not be as effective in roundabouts’ crosswalks with multiple lanes at an entrance or exit, where drivers may not yield because they may not see the pedestrian when other yielding vehicles block their view. Gazing behavior did not result in a statistically significant change in drivers’ behaviors. This finding confirms previously reported results at signalized crosswalks (Bourquin, Wall Emerson, Sauerburger, & Barlow, 2016). Gazing alone had very low yielding rates; when gazing was added to other conditions, it did not result in any consistent change in yielding. Higher driver yielding in general mitigates the effects of errors in judgement by pedestrians. Furthermore, detectable drivers’ yields can be an essential component, along with other factors, which may be used by pedestrians who are visually impaired when determining if and when to initiate a crossing at a roundabout. O&M specialists should consider a curriculum for crossing at roundabouts that incorporates the hand up and reversible step techniques.
LIMITATIONS OF THE STUDY
The sample size for this study was relatively small. It was conducted at a single roundabout, so the authors cannot report with certainty on how generalizable the results may be or how they may be applied at other locations. We did not explore many other possible roundabout configurations.
This study used one male subject for all trials; it did not examine the possible interactions of gender and yielding.
These behaviors were explored before or at the start of a street crossing; we do not know if the findings apply to using any of these behaviors such as head or eye movements or raising the hand in the middle of the crossing when the threat from a vehicle is in the second half of the crosswalk with vehicles at a proscribed distance from the crosswalk.
The authors note again that the pedestrian has the right-of-way to enter the crosswalk, and that drivers are required to yield. Yields, however, may raise the risk of a rear-end collision resulting from a vehicle that is too closely following the one yielding (we did not, however, collect any data related to this issue). Also, one driver yielding in one lane of a multilane crosswalk does not mean that other approaching drivers will also yield. The effects on drivers at roundabout multiple-lane exits and entrances were not fully examined. O&M specialists and pedestrians who are visually impaired must also consider those variables found to be relevant in previous studies: the speed of approaching vehicles, whether vehicles are closely following each other, and the number of lanes to be crossed. Therefore, the authors recommend caution when applying these findings at multi-lane crossings at roundabouts.
References
- Ashmead, D. H., Guth, D., Wall, R. S., Long, R. G., & Ponchillia, P. E. (2005). Street crossing by sighted and blind pedestrians at a modern roundabout. Journal of Transportation Engineering, 131(11), 812–821.
- Barlow, J. M., Bentzen, B. L., Sauerburger, D., & Franck, L. (2010). Teaching travel at complex intersections. In R. L. Welsh & B. B. Blasch (Eds.), Foundations of orientation and mobility (third ed., Vol. II, pp. 352– 419). New York, NY: American Foundation for the Blind.
- Bourquin, E., Wall Emerson, R., Sauerburger, D., & Barlow, J. (2014). Conditions that influence drivers’ yielding behavior in turning vehicles at intersections with traffic signal controls. Journal of Visual Impairment & Blindness, 108(3), 173–186.
- Bourquin, E., Wall Emerson, R., Sauerburger, D., & Barlow, J. (2016). Conditions that influence drivers’ yielding behavior: Effects of pedestrian gaze and head movements. International Journal of Orientation and Mobility, 8(1), 13–26.
- Elvik, R. (2003). Effects on road safety of converting intersections to roundabouts: A review of evidence from non-US studies. Journal of the Transportation Research Board, 1847, 1–10.
- Geruschat, D. R., Fujiwara, K., & Wall Emerson, R. S. (2011). Traffic gap detection for pedestrians with low vision. American Academy of Optometry, 88(2), 208 –216.
- Geruschat, D. R., & Hassan, S. E. (2005). Driver behavior in yielding to sighted and blind pedestrians at roundabouts. Journal of Visual Impairment & Blindness, 99(5), 286 –302.
- Geruschat, D. R., Hassan, S. E., & Turano, K. A. (2003). Gaze behavior while crossing complex intersections. Optometry and Vision Science, 80(7), 515–528.
- Guth, D., Ashmead, D., Long, R., Wall, R., & Ponchillia, P. (2005). Blind and sighted pedestrians’ judgments of gaps in traffic at roundabouts. Human Factors, 47(2), 314 – 331.
- Hassan, S. E., Geruschat, D., & Turano, K. A. (2005). Head movements while crossing streets: Effect of vision impairment. Optometry & Vision Science, 82(1), 18–26.
- Inman, V. Davis, G., & Sauerburger, D. (2006). Pedestrian access to roundabouts: Assessment of motorists’ yielding to visually impaired pedestrians and potential treatments to improve access. Washington, DC: Federal Highway Administration. Retrieved from http://www.fhwa.dot.gov/ publications/research/safety/pedbike/05080/ 05080.pdf
- Kalamazoo Area Transportation Study. (2011). Traffic control database system. Retrieved from http://katsmpo.ms2soft.com/tcds/tsearch. asp?loc=Katsmpo&mod=
- Long, R. G., Guth, D. A., Ashmead, D. H., Wall Emerson, R., & Ponchillia, P. E. (2005). Modern roundabouts: Access by pedestrians who are blind. Journal of Visual Impairment & Blindness, 99(10), 611–621.
- Man, S., Cho, D. S., Choi, T. M. (2007). Influence of pavement surface noise: The Korea Highway Corporation test road. Canadian Journal of Civil Engineering. 34(7), 809 –816.
- Schroeder, B., Hughes, R., Rouphail, N., Cunningham, C., Salamati, K., & Long, R. (2011). NCHRP report 674 crossing solutions at roundabouts and channelized turn lanes for pedestrians with vision disabilities. Washington, DC: National Cooperative Highway.
- Wall Emerson, R., Naghshineh, K., Hapeman, J., & Wiener, W. (2011). A pilot study of pedestrians with visual impairments detecting traffic gaps and surges containing hybrid vehicles. Transportation Research Part F 14, 117–127.
- Wall Emerson, R., & Sauerburger, D. (2008). Detecting approaching vehicles at streets with no traffic control. Journal of Visual Impairment & Blindness, 102(12), 747–760.
Eugene A. Bourquin, D.H.A., orientation and mobility specialist and owner, Bourquin Consulting, 315 8th Avenue, Apartment 10E, New York, NY 10001; e-mail: oandmhk@msn.com.
Robert Wall Emerson, Ph.D., professor, Department of Blindness and Low Vision Studies, Western Michigan University, 1903 West Michigan Avenue, Kalamazoo, MI 49008; e-mail: robert.wall@wmich.edu.
Dona Sauerburger, M.A., orientation and mobility specialist (independent practitioner); mailing address: 1606 Huntcliff Way, Gambrills, MD 21054; e-mail: dona@sauerburger.org.
Janet M. Barlow, M.Ed., orientation and mobility specialist and president, Accessible Design for the Blind, 3 Manila Street, Asheville, NC 28806; e-mail: jmbarlow@accessforblind. org.